TURBOMOLE Users Forum
TURBOMOLE Modules => Mpgrad and Rimp2 => Topic started by: lyzhu on June 11, 2009, 10:31:27 PM
-
Hi,
I am trying to do CP-corrected optimization using rimp2 method of the CC2 module in Turbomole6.0. Firstly, I used the jobbsse -setup option to generate *_ghosts, monomer_* and supermol subdirectory, then I checked numbers of frozen orbital in the *_ghosts subdirectory. Then I do "jobbsse -ri -level cc2 -c 300 -opt >jobex.out" calculation. But there is a error occured:
*******************************************************
Error in DFT/SCF step... please check:supermol/dscf.out.supermol
========================
internal module stack:
------------------------
bsseenergy
========================
DFT/SCF did not converge!!! Please check!
bsseenergy ended abnormally
***********************************
Then I checked the supermol/dscf.out.supermol file, it is:
STARTING dscf ON 8 PROCESSORS!
RUNNING PROGRAM /export/apps/TURBOMOLE60/bin/em64t-unknown-linux-gnu_mpi/dscf_mpi.
PLEASE WAIT UNTIL dscf HAS FINISHED.
Look for the output in slave1.output.
NOTE: the number of nodes in your machine list:
125091_hosts
is LOWER than the number of nodes requested
PARNODES has been set to 8
PARNODES will be ignored - change your machine file
to use more nodes. Remember to add a node with several CPUs
multiple times to your machine file (one line = one CPU)
Calculation will continue on 0 CPUs
MACHINEFILE is 125091_hosts
No file slave1.output found?
****************************************
But when I checked the machine file , it included 8 nodes.
What's the matter? Could anyone give some suggestions?
Thanks!
-
Hi,
the jobbsse script is not prepared for parallel calculations. For this purpose it would be neccessary to extend it such that
it keeps properly track of the statistics and task distribution settings for each of the sub cases (ghosts, monomer, supermol).
Christof